None Surviver_Predictions

The Titanic

titanic-dock.jpg

Survivor Predictions

                    Utilizing Ensemble Machine Learning Techniques
                            by: John-Eric Bonilla, MSDA

Table of Contents

Goals and Objectives

Project Overview

Import Dependancies

Read in and view dataset

Data Cleaning - Continuous Features

The Age Attribute

The 'SibSp' and 'Parch' Attributes (Siblings & Spouse)

Data Cleaning - Categorical Features

Visualize Cleaned Dataset

Partition dataset

* Train, test, and validation split

* Validate partition size.
* Save partitioned datasets

Ensemble 'Boosting' Model

Hyperparameter Tuning

Save boosting model

Ensemble 'Bagging' Model

Hyperparameter Tuning

Save bagging model

Ensemble 'Stacking' Model

Save stacked model

Comparing the different ensemble models

Validation Set Evaluation

Assessing the best performing model (gradient-boosting) on the test data.

Conclusion:

    *  The GB model perfomed even better on the test set then the validation set.